First, we had classified the firms which issued ESOs by industry type. Totally 1099 firms selected and it’s including TSE and OTC market, sample period refers from 2001 to 2010.
Figure 4- 1: The chart-pie of firm insustry classification on issuance of ESOs
Issuance
18
Above figure presents that the high-tech industries which the leading sectors of Taiwan are more likely to issue ESOs for their compensation, including semiconductor, electronics and computer industry firms. In many cases in Taiwan, stock compensation represent a very substantial portion of an employee’s total annual compensation in high-tech industry, unlike in the US where stock ownership may be only a small in the total compensation of an employee. The explanation for this is that the salary level in Taiwan’s high-tech sector is relatively lower than that in subsidiaries of large US multinational corporations, like IBM and HP. Hence, many high-tech firms in Taiwan adopt this sharing scheme to make their total compensation package more competitive in the market, thereby putting them on equal footing in order to attract and retain better personnel.
Figure 4- 2: Time trend for number of TSE and OTC firms issued ESOs
In 2001, the Financial Supervisory Commission (FSC) of Taiwan amended the Securities and Exchange Act and since then allowed public firms to grant stock options to their top executives and employees.
From 2001 to 2003 the issuance increased, but from 2003 to 2005 it’s declined because of might be the bad performance of securities market, it affects the stock options’
price performance, and it could cause of why the warrants cannot be the enough motivation.
Another reason was the financial accounting standards’ changing that the Accounting Research and Development Foundation (ARDF) of Taiwan issued Interpretations No.70, 71, 72 and 205, requiring firms to use the fair value method or the intrinsic value method to account for their ESOs. Beginning 2004, if choosing to use the fair value method, firms should first measure the fair value of options expense over the vesting period. Alternatively, if firms choose to use the intrinsic value method, they need to disclose the ESO-related expense information in the footnotes to the financial statements as if firms had employed the
0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
TSE OTC Total
19
fair value method to price their ESOs. In fact, almost all firms choose to use the intrinsic value method; and thus typically recognize nothing in their income statements.
In 2007 August, the ARDF of Taiwan issued Statement No.39, Accounting for Stock-Based Compensation, which requires firms to use the fair value method to account for options granted on or after January 1, 2008. Under No.39 firms do not have to apply retrospective application for options granted prior to 2008; however, they still need to disclose the option-related pro forma information in the footnote for options granted prior to 2008; and meanwhile, recognize ESO-related compensation expense for options granted in 2008. As a result, the shares and warrants that were widely issued by local corporations as forms of employee profit sharing are now required to be market-to-market, with differences recognized through the income statement. For this reason, many companies have applied to the FSC for the issuance of employee stock options prior to the effective date of SFAS No39 (You Quan, Lin 2009). Please see the figure 4.2, the number of ESO-issuing firms spurted in 2007 and went down in 2008.
20
Table 4- 1: Classification of ESO issued firms by industry type
No Industry type 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Total
% TSE OTC TSE OTC TSE OTC TSE OTC TSE OTC TSE OTC TSE OTC TSE OTC TSE OTC TSE OTC TSE OTC 1 Semiconductor 4
9
9 Chemical, Biotechnology and medical
care 1 3 2 3 2 11 2%
Note: 547 high-tech industry firms with 93% of total sample data which including semiconductor, electronic components, optoelectronic, computer and peripheral equipment, communication and internet, electronics product distribution, electronic machinery, information service and other electronic industry had issue the ESOs in last ten years.
The 599 firms with 54,5% is Taiwan Stock Exchange market (TSE) firms and 500 firms with 45,5% is Over the counter market (OTC) firms.
21
4.2 Descriptive statistics and correlations
After excluding firms with incomplete data, 14114 firm-years remained, composing 964 firm-years with options granted and 13150 firm-years without option grants. Follow Table 4.2 present descriptive statistics on the variables that provides the means, median, standard deviations, minimum, maximum, first and third quartiles. Table 4.3 shows the correlations between variables.
Table 4- 2: Descriptive analysis of each variable
ROA FCF ratio Tobin's Q Sales Per Empl Dir's % Manag's' % Out 10's %
Total 8,315 0,839 0,821 13161,198 28,354 5,973 17,662
0 8,042 0,859 0,780 13145,811 28,691 5,986 17,877
1 12,044 0,566 1,373 13371,083 23,769 5,792 14,727
Total 7,835 1,990 0,592 6604,500 24,020 3,115 16,290
0 7,620 1,991 0,566 6566,500 24,305 3,080 16,540
1 11,885 1,972 1,042 7356,500 20,895 3,510 13,805
Total 11,749 15,540 0,928 22194,816 18,318 7,674 12,013
0 11,652 15,623 0,890 22364,098 18,568 7,768 12,189
1 12,424 14,352 1,209 19750,451 13,693 6,245 8,795
Total ‐87,800 ‐233,710 0,000 ‐6612,000 0,000 0,000 0,000
0 ‐87,800 ‐233,710 0,000 ‐6612,000 0,000 0,000 0,000
1 ‐73,080 ‐84,173 0,000 108,000 2,010 0,000 0,000
Total 114,620 831,036 16,471 564438,000 100,000 76,760 91,140 0 114,620 831,036 16,471 564438,000 100,000 76,760 91,140
1 55,380 54,851 11,392 265993,000 89,220 45,890 48,700
Total 2,570 ‐3,944 0,267 3632,250 15,620 0,560 9,333
0 2,410 ‐3,766 0,248 3593,000 15,790 0,520 9,420
1 5,598 ‐6,351 0,630 4141,500 13,815 1,370 8,590
Total 14,398 7,583 1,087 13562,250 36,258 8,430 24,218
0 13,970 7,503 1,034 13505,000 36,760 8,460 24,640
1 19,000 9,046 1,742 14175,000 30,230 8,280 19,870
Total
Note: “Total” rows show the total number of sample, “0” rows present didn’t issue ESOs’ firm-year and “1” rows present issued ESOs’ firm-year which about 7,33% of total firms had issuance.
In the Table 4.2, the mean of ROA between ESO issued firms was 12,044, it’s higher than 8,315 in the total samples. The mean of FCF ratio in the ESO issued firms was 0,566 while it was 0,839 in the total samples. The mean of Tobin’s Q in the ESO issued firms was 1,373, it’s much higher than 0,78 in the did not issue ESO firms while it was 0,821 in the total sample.
From here, we can make a simple analyze on firm-specific characteristics that the firms with higher ROA and Tobin’s Q, lower FCF ratio are more likely to issue ESOs to their compensation.
22
From the mean of Sales per employee and Managers’ share percent we could not see any specific, but on the mean of Board of directors’ share and Top outside 10 shareholders’ share percent, we can see that these two variables have negatively relationship with Issuance of ESOs.
The failure of hypothesis 4 may arises the various type of industry in this study sample.
Table 4- 3: Correlations between variables
ESO ROA FCF ratio Tobin's Q Sales Per Empl Dir's' % Manags' % Top out 10 %
ESO 1
ROA 0,086 1
FCF ratio ‐0,005 0,307 1
Tobin's Q 0,161 0,428 0,100 1
Sales Per Empl 0,003 0,043 0,049 0,026 1
Dirs' % ‐0,068 0,055 0,014 ‐0,081 ‐0,013 1
Manags' % ‐0,006 0,102 0,032 0,084 ‐0,063 0,076 1
Top out 10 % ‐0,066 ‐0,054 0,051 0,092 0,084 ‐0,263 0,009 1
Note:
4.3 Testing hypothesizes
Table 4.5 present the logistic regression on each variables. Asterisks ***, ** and * denote statistical significant at 0.5%, 5% and 10% levels, respectively.
Variables in the Equation
B S.E. Wald df Sig. Exp(B)
Step 1a ROACEBIDA .012 .003 12.379 1 .000 1.012
FCFTotalasset -.006 .003 6.412 1 .011 .994
TobinsQ .397 .031 160.443 1 .000 1.488
SalesPerEmployee .000 .000 .349 1 .555 1.000
Directorsshare -.023 .002 91.132 1 .000 .978
Managersshare -.007 .005 1.738 1 .187 .993
Topout10 -.036 .003 108.595 1 .000 .965
Constant -1.926 .096 402.046 1 .000 .146
23
Table 4- 4: Logistic regression coefficients
Independent Factors Wald Satistics (sig.) 12.379
(0.000) Dependent Factor: "Issuance of ESOs"
0.032 0.083 6570.028a
Note: *** p < 0.005, ** p< 0.05, * p< 0.1 Note:
From the result we can write the logistic regression equation as follow:
Logit(p)= -1.926 + 0.012*ROA - 0.006*FCF + 0.397*Tobin’s Q - 0.023Dir.% - 0.036Top.10%
The Table 4.5 provides, Return on Asset (ROA, β1=0,012, p value<0.005) and Growth Opportunity (Tobin’s Q, β3=0,397, p value<0.005) are have positive relationship with Issuance of ESOs, thus hypothesizes 1 and 3 are supported.
Free cash flow (FCF, β2=-0.006, p value<0.05); Board of Directors’ share percent in total share (Directors’ share, β5=-0.023; p value<0.005) and Top outside 10 shareholders’ share percent on total share (Top out 10, β7=-0.036, p value<0.005) are have negative relationship with Issuance of ESOs, thus hypothesizes 2, 5 and 7 are supported.
Unfortunately, in our regression result the Sales per Employee (β4=0.000, p value>0.1) and Managers’ share percent on total shares (β6=-0.007, p value>0.1) doesn’t have any relationship with Issuance of ESOs and thus hypothesizes 4 and 6 are did not supported.
24
CHAPTER 5: CONCLUSIONS